“We need to use artificial intelligence” is an intention, but it is not yet a project. To start, the company needs to identify a specific challenge and define how it will know when it has been resolved.
Microsoft’s October 2026 announcements highlight new features for CRM, customer service, and business process agents, at various stages of availability. This development provides context for a question closer to day-to-day operations: Which task deserves to be automated first?
Choose an observable problem
Our recommendation is to start with repetitive activities that the team understands well. Describe the current workflow before choosing the technology: what triggers the task, what information is needed, who is involved, and what indicates completion.
For example, preparing an internal summary of requests can be a more focused pilot project than handing over all customer communication to a system.
Check if the process is well-defined
If two people perform the same task in incompatible ways, find out why. Perhaps there are legitimate exceptions; perhaps criteria are missing. Automating before clarifying these differences may simply replicate the confusion.
Also identify incomplete information. A tool does not eliminate the need to keep data understandable and accessible to those who need to use it.
Conduct a pilot with limits
Define which inputs will be used, who will review the results, and under what circumstances the workflow should stop. Include challenging examples in the evaluation, rather than testing only the simplest cases.
During the pilot, track the time spent, the number of corrections, and the usefulness of the results. Compare these with the previous execution. If the review requires more effort than the original process, investigate the cause before scaling up.
Make decisions based on what you’ve learned
A test may indicate that the best solution is a simple integration, an improved form, or a conventional rule—without AI. It may also reveal a valuable application that deserves further development.
The goal is to improve work. The technology chosen should support that decision.
Virealto helps transform business needs into digital solutions by combining process analysis, development, and the guided use of artificial intelligence.
